AI agents & systems · Taking projects
We run on
We run on
our own agents.
This studio is run day to day with AI agents working alongside the founder: planning, research, writing, code, testing, publishing. The agents do the work and report; a human decides.
What we really do
- Daily operationsThis website, the products in our line and the RAMAS artist project were planned, written, coded, tested and published with AI agents working alongside the founder.
- How decisions workAgents prepare, draft and check; a person approves what goes out. Nothing is published, sent or deleted without a human decision.
- What we build for youAn assistant for one concrete workflow — research, drafting, reporting, support replies, content pipelines — connected to the tools you already use.
- Limits, stated up frontWe say what an agent is good at, where a person still has to decide, and what it must never do. That boundary is part of the delivery.
- Scope and timingOne or two agents for a specific workflow, delivered in weeks, with a written description of what they may and may not do. Not a platform.
We don’t promise autonomy that doesn’t exist. If a task needs a person, we tell you so before we start.
Have a workflow that eats your week?
Start a project